A feature of modern democracies is public mistrust of scientists and the politicization of science policy, e.g., concerning stem cell research and genetically modified food. While the extent of this mistrust is debatable, its political influence is tangible. Hence, science policy researchers and science policy makers need early warning of issues that resonate with a wide public so that they can make timely and informed decisions. In this article, a semi-automatic method for identifying significant public science-related concerns from a corpus of Internet-based RSS (Really Simple Syndication) feeds is described and shown to be an improvement on a previous similar system because of the introduction of feedbased aggregation. In addition, both the RSS corpus and the concept of public science-related fears are deconstructed, revealing hidden complexity. This article also provides evidence that genetically modified organisms and stem cell research were the two major policyrelevant science concern issues, although mobile phone radiation and software security also generated significant interest.
IntroductionPublic mistrust of science has probably existed as long as science itself. Historical manifestations have varied from popular culture such as Mary Shelley's Frankenstein and H. G. Wells' The Island of Doctor Moreau (Pinsky, 2003;Wilt, 2003;Wolpert, 2005), to mass antinuclear power movements (Herring, 2006;Hsu, 2005). In recent years, many scientific issues have become politicised and have given rise to pressure groups, media coverage, and public debates. Two of the most prominent have been stem cell research and genetically modified food, both of which have triggered significant social sciences and ethics research (Hagendijk, 2004;Tait, 2001;Tsai, 2005). The politicization of science debates has lead to government policy and legislation curtailing researchers' activities in response to public pressure. The importance of this is potentially great (Leydesdorff & Etzkowitz, 2003). Given the key role of research within a profitable modern knowledge-based economy (Etzkowitz & Leydesdorff, 1997;Gibbons, 1994), for example, in the biotechnology and computing industries, falling behind with a newly emerging technology is a national potential disaster. Conversely, allowing science to continue "unchecked" (except for the normal self-regulation process) may cost humanity too high a price if the critics are correct (Chadwick, 2005;London, 2005). In consequence, a large amount of research has been devoted to topics such as the sociology and ethics of individual science policy debates (Klotzko, 2004). It is hence particularly important to be able to identify the next significant science concern debate as early as possible so that research into the social, ethical, legal, and policy implications can begin and support politicians to make wellinformed, timely decisions. The research reported here is part of an international European Union-funded project (www.creen.org) that aims to develop automatic methods to help identify critical p...